Coral Bleaching as an Adaptive Mechanism

Author(s): Robert W. Buddemeier and Daphne G. Fautin

Source: BioScience, Vol. 43, No. 5 (May, 1993), pp. 320-326
Published by: University of California Press on behalf of the American Institute of Biological Sciences
Stable URL: http://www.jstor.org/stable/1312064 .
Accessed: 15/09/2013 02:28

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

University of California Press and American Institute of Biological Sciences are collaborating with JSTOR to
digitize, preserve and extend access to BioScience.

http://www.jstor.org

This content downloaded from 66.194.72.152 on Sun, 15 Sep 2013 02:28:36 AM

All use subject to JSTOR Terms and Conditions
 Coral Bleachingas an
Adaptive Mechanism
Robert W. Buddemeier and Daphne G. Fautin
leaching in organisms such as
hard and soft corals, giant
clams, and sea anemones is the
loss of pigment associated with their
symbiotic algae (zooxanthellae).
Bleaching can be the result of loss of
the pigmented algae from the host or
of loss of pigment from the algae. This
article considers only the phenom-
enon of algal loss. Most attention has
been directed to bleaching in
scleractinian (hard) corals in the trop-
ics (e.g., Brown 1990, Glynn 1990,
Williams and Bunkley-Williams
1990). It is a subject of concern be-
cause of mortality and local extinc-
tions associated with large-scale
bleaching episodes (Glynn and de
Weert 1991; but see Glynn and Fein-
gold 1992), the widespread belief that
bleaching episodes have increased dra-
matically in frequency and inten-
sity within the past decade (D'Elia et
al. 1991), and speculation about pos-
sible links to global warming (Glynn
RobertW. Buddemeieris a senior scien-
tist and chief of the GeohydrologySec-
tion at the Kansas Geological Survey,
Lawrence,KS 66047, and courtesypro-
fessor of geographyat the Universityof
Kansas,Lawrence,KS 66045. His inter-
ests include coral growth, reef and atoll
environments,hydrologic systems, and
the effectsof climatechangeand variabil-
ity. Daphne G. Fautinis an adjunctpro-
fessor in the Departmentof Systematics
and Ecology, Snow Museum, University
of Kansas,Lawrence,KS 66045, and an
adjunct senior scientist at the Kansas
GeologicalSurvey,Lawrence,KS 66047.
She is a coelenterate biologist with a
long-standingresearchinterestin marine
symbioses.? 1993 AmericanInstituteof
Biological Sciences.
320
This content downloaded from 66.194.72.152 on Sun, 15 Sep 2013 02:28:36 AM
A testable hypothesis
We posit that bleaching
allows a host to be
repopulatedwith a
differentpartner
1991, Goreau 1990, Williams and
Bunkley-Williams 1990).
In this article, we examine briefly
some of the salient attributes of bleach-
ing and its biological participants, to
which we refer by the names used in
the primary literature. We fully real-
ize that these names may not be accu-
rate (see Knowlton et al. 1992 for an
example) and that taxonomic preci-
sion of both partners is essential for
understanding symbioses. We propose
a conceptual hypothesis for the role of
bleaching in coral reef ecology and
evolution that is consistent with most
of the available data. We indicate
how this hypothesis can be tested, and
we discuss its implications for research,
management, and conservation.
Acute bleaching is a generalized
stress response resulting from a vari-
ety of environmental conditions out-
side the normal local range. It can be
induced by stimuli such as unusually
high or low temperatures, low light
levels or constant light, high levels of
UVradiation, salinity shock, and other
chemical insults (summarized by, e.g.,
Cook et al. 1990, D'Elia et al. 1991,
Gates 1990), some or all of which may
operate synergistically (Coles and
Jokiel 1978). Much recent concern
has focused on the result of high-
All use subject to JSTOR Terms and Conditions
temperature stress associated with ei-
ther the 1982-1983 El Nifio event or
with the 1987 Caribbean warm event.
However, partial or low-level bleach-
ing also occurs in the absence of per-
ceptible stress (Fisk and Done 1985,
Gates 1990). Such background bleach-
ing may well have been regarded as an
unexceptional feature of coral reef
variability before the concern gener-
ated by mass bleachings and associ-
ated mortality in the middle to latter
1980s (Oliver 1985). Thus, bleaching
appears to be a basic physiological
attribute of many, if not all, organ-
isms having zooxanthellae, both in
response to a variety of stresses and in
the absence of obvious stress.
Characteristics of bleaching
Based on published data, we identify
three key features of coral bleaching
attributed to high temperature:
* The upper temperature limit (be-
yond which bleaching results) is lo-
cally defined: corals may experience
local maxima from the mid-20s to the
mid-30s degrees C, but at least some
corals of most taxa bleach when the
temperature rises a few degrees above
the local average maximum (Coles et
al. 1976, Cook et al. 1990, Jokiel and
Coles 1990). Temperature sensitivity
is therefore not specific to the coral
taxon, but it involves some other com-
ponent of adaptation.
* There are consistent habitat differ-
ences in bleaching resistance at a given
locale: corals in habitats that are more
variable or more prone to stresses of
various sorts, including thermal (such
BioScience Vol. 43 No. 5
as shallow water or fringing reefs),
tend to be less bleached than those in
more normally equable environments
(Cook et al. 1990, Hoeksema 1991).
* Despite apparent environmental ad-
aptations, there are consistent taxo-
nomic differences in vulnerabilities to
stress-related bleaching and in associ-
ated mortality at a particular site (e.g.,
Cook et al. 1990, Hoeksema 1991,
Jaap 1985). Prevalence of bleaching
appears to be higher in species that
have high rates of recruitment and
grow rapidly (weedy species) than in
other corals in the same general envi-
ronment (e.g., Gates 1990, Jokiel and
Coles 1990, Salvat 1991), and mortal-
ity after bleaching is higher in some
branching species-particularly mem-
bers of Acropora-than in many mas-
sive corals (e.g., Brown and Suharsono
1990, Fisk and Done 1985, Oliver
1985, Salvat 1991). However, intra-
specific variability in environmental
sensitivity has also been documented,
with some but not all sympatric colo-
nies/individuals bleaching (e.g., Brown
and Suharsono 1990, Hoeksema 1991,
Oliver 1985). Jokiel and Coles (1990)
identified temperature-sensitive and
temperature-resistantclones of Porites
compressa in Hawaii.
The underlying assumption in studies
of algal symbiosis formerly was that
all zooxanthellae belong to Symbio-
dinium microadriaticum Freudenthal,
1962. The evidence for diversity is
now overwhelming: biochemical
(Ciereszko 1962, 1991, Trench 1971,
Von Holt and Von Holt 1968), elec-
trophoretic (Schoenberg and Trench
1980a), morphological (e.g., Blank
1987, Blank and Trench 1985a), and
genomic (Rowan and Powers 1991).
Whether the algal populations for-
mally are species, races, strains, or
ecotypes, their functional distinguish-
ability is adequate to support the hy-
pothesis we advance. We refer to them
as types. We recognize that our hy-
pothesis would be invalidated by the
finding of only one possible combina-
tion of host species and algal type as
surely as it would be by the existence
of a single type of alga in all hosts.
The hypothesis
From the available data on the plant-
animal symbiosis, and by analogy with
other symbioses (e.g., Fautin 1991) in
May 1993
This content downloaded from 66.194.72.152 on Sun, 15 Sep 2013 02:28:36 AM
Corals in close proximity such as these Malaysian Turbinaria(left foreground)and
Acropora (background)may host differenttypes of algae. Photo: D. G. Fautin.
which partners range from generalists
to specialists, we consider it likely
that some taxa of hosts and algae can
form stable associations with more
than one taxon of symbiont. On that
basis, we propose that bleaching has
the following attributes:
* Physiological responses (including
those related to stress) of the symbi-
otic unit are primarily characteristics
of the combination, rather than of the
host or symbiotic algal partner alone.
* Symbiotic units with one common
partner vary in their stress responses
and the degree to which they are
adapted to different environmental
conditions.
* Multiple types of both zooxanthel-
lae and host species are commonly
available on reefs.
* Bleaching provides an opportunity
for the host to be repopulated with a
different type of partner; frequent
stress tends to favor a stress-resistant
combination.
* Altered environmental conditions
(which include altered frequency of
stressful events and altered types of
stress) favor establishment of combi-
nations of symbionts that were less
adaptive under previous conditions.
* Stress-sensitive combinations may
have competitive advantages in the
absence of stress, which implies
All use subject to JSTOR Terms and Conditions
gradual reversion to stress-prone but
otherwise advantageous symbiotic
units in stable environments or in the
absence of extreme conditions.
Host-alga symbioses
Symbiont specificity. Uniform, wide-
spread bleaching is uncommon, even
among members of a single host spe-
cies in a particular locale under wide-
spread stress (e.g., Brown and
Suharsono 1990, Lang et al. 1988,
Salvat 1991), and commonly bleach-
ing occurs only in portions of some
colonies (Hayes and Bush 1990, Oliver
1985) or even individuals (Hoeksema
1991). We suggest that at least some
of this variation is due to different
types of algae in conspecific hosts or
even in parts of a genetically uniform
host unit.
Schoenberg and Trench (1976,
1980a) found that some species of
hosts contain more than one type of
zooxanthella; in some cases the host
units were sympatric, in others they
were geographically separated-but a
single individual or colony had algae
of only one type. Rowan and Powers
(1991) found two restriction fragment
length polymorphisms of single-
stranded small ribosomal subunit RNA
in algae from one of the ten species of
reef-forming corals for which mul-
tiple samples were examined (as well
321
as in a congeneric species, details for
which were not provided), although
Rowan1 doubts that that result is taxo-
nomically significant. Morphological
and physiological differences in algae
from well-lighted and shaded parts of
a cnidarian colony (e.g., Berner et al.
1987 and studies cited therein) may
imply multiple types of zooxanthel-
lae, contrary to the usual interpreta-
tion of photoadaptation, by analogy
with experiments on monocultures of
free-living phytoplankters. Level of
illumination may be important in de-
fining microhabitats suitable for dif-
ferent types of algae, and thus account
for bleachingin corals moved into unfa-
miliar light regimes (Dustan 1982).
The potential for exchange of sym-
bionts has been demonstrated experi-
mentally: Kinzie and Chee (1979)
infected a sea anemone with zoo-
xanthellae from a clam, a nudibranch,
and a medusa; Schoenberg and Trench
(1980a) infected an anemone with
algae from another anemone, a me-
dusa, and three corals; and Fitt (1985)
induced production of medusae in
scyphozoan polyps with zooxanthel-
lae from 19 other species, including
anemones, corals, zoanthids, and
clams, and he infected larval clams of
two species with algae from anemo-
nes and zoanthids. In the field,
Schoenberg and Trench (1980a) and
Rowan and Powers (1991) found al-
gae of a particular type in hosts be-
longing to different classes or phyla.
Although many hosts and algae can
live with more than one symbiotic
partner, there is a spectrum of their
abilities to do so. Some types of algae
will not establish in heterologous hosts
(Schoenberg and Trench 1980b), oth-
ers establish but do not persist (Kinzie
and Chee 1979), and some are a poor
physiological fit, as demonstrated by
their failure to induce medusa pro-
duction in scyphozoan polyps (Fitt
1985). Specialists may be more obli-
gate than generalists, the degree of
bleaching mortality possibly reflect-
ing the dependence of hosts on their
symbionts. For example, the virtually
universal mortality of bleached corals
of the genus Acropora (Salvat 1991)
suggests a tightly bound combination.
The range of specificity and stabil-
1R. Rowan, 1992, personal communication.
Australian Institute of Marine Science,
Townsville, Queensland, Australia.
322
This content downloaded from 66.194.72.152 on Sun, 15 Sep 2013 02:28:36 AM
ity among partners provides variabil-
ity on which natural selection can
operate and the potential for recombi-
nation in the event of mass stress,
although we recognize that for a zoox-
anthella to have one primary pre-
ferred host (or even a few) would be a
stabilizing community force in the
short term. Repopulation following
bleaching may result in any of the
following: an unstable relationship
that breaks down rapidly; a stable
relationship in which the alga may be
eliminated by competitively superior
types over evolutionary/population
time scales of the host; an advanta-
geous combination resistant to inva-
sion by other symbionts under
nonstress conditions; or coexistence
of more than one type, as in the sea
anemones Anthopleura elegantissima
and Anthopleura xanthogrammica,
which may simultaneously harbor
zoochlorellae and zooxanthellae
(Muscatine 1971).
Because a symbiotic unit is likely to
have attributes possessed by neither
partner in isolation (Karakashian and
Siegel 1965), we expect a species of
coral to vary in stress response with
symbiont type and a type of zooxan-
thella to vary in physiology depending
on its host. For any physiological prop-
erty, assessing the contribution of one
symbiont is not a simple matter of
subtracting from the value for the
intact symbiosis the contribution of
the other symbiont measured in isola-
tion. Experiments by Trench (1971),
for example, provided preliminary
evidence that a particular type of zoo-
xanthella differs biochemically in host
corals of various species, and those by
Schoenberg and Trench (1976,1980b)
demonstrated that algae of various
types differ when grown in genetically
identical hosts.
In addition to physiological fea-
tures of the partners and the combina-
tion, we hypothesize that environ-
mental conditions affect specificity.
This idea is not new: Kinzie and Chee
(1979) pondered whether ecological
setting might be more important in
explaining features of the symbiosis
than identity of the partners. A return
to prestress conditions after bleaching
may favor restoration of the original
symbiotic combination. However,
when the stress represents a shift in
environmental conditions, the combi-
nation is likely to change, if for no
All use subject to JSTOR Terms and Conditions
other reason than the pre-existing com-
bination experienced those conditions
as stressful. Laboratory studies dem-
onstrating that homologous algae dis-
place experimentally introduced het-
erologous ones (Fitt 1985) illustrate
precisely this characteristic, for the
experimental setting is manipulated
to approximate favorable natural con-
ditions, which are those of the pre-
existing symbiotic unit. Frequency of
episodic stress will also affect which
combination (or combinations) per-
sists, because reversion is rarely in-
stantaneous.
Bleaching mechanism. How bleach-
ing occurs is poorly understood. From
laboratory experiments on a coral and
a sea anemone exposed to low-tem-
perature shock, Gates et al. (1992)
inferred a breakdown in host-cell ad-
hesion. It is unknown how general
this mechanism is across taxa of hosts
and types of algae, or how it relates to
duration and intensity of stress. Com-
plete sloughing of the endodermal cells
in which zooxanthellae lodge seems
to us pathological, and therefore un-
likely to account for the reversible
bleaching that appears characteristic
of many corals. We believe that in
some hosts or under some stresses, the
mechanism of loss/regulation is prob-
ably less violent and the mechanism of
Gates and colleagues may operate
primarily in taxa with high mortality
rates attendant upon bleaching or in
cases of severe stress. Alternatively,
some host species may be more toler-
ant of such damage. Our hypothesis
does not depend on a particularmecha-
nism nor on a determination of
whether the host (as is commonly
assumed) or algal partner initiates
bleaching.
Mechanisms of infection and algal
spread. Our hypothesis requires that
free-living zooxanthellae (swarmers)
be available as colonizers. They pre-
sumably must be conveyed through
water, although, given that densities
in seawater are low, vectors such as
crustaceans or fish may be important
(Fitt 1984, Muller Parker 1984). De-
tailed predictions based on our hy-
pothesis will differ depending on
whether the process of bleaching ren-
ders either partner temporarily or per-
manently unable to form a symbiosis,
and whether swarmers are affected by
BioScience Vol. 43 No. 5
stress similarly to the same type ofbut not to symbiotic hosts (Fitt 1984). if intervals between bleaching events
alga in a symbiotic unit. Factors determining which types of are long compared with the time re-
algae and how many of each survive
We infer that the repetitive nature quired for reversion, and if the envi-
of small-scale bleachings (Jokiel and
may include, in addition to external ronment returns to its prestress state.
Coles 1990) reflects a lack of reas-environment, physiological compat- The resulting advantageous but vul-
ibility, initial numbers, and growth
sortment in symbiotic partners. If the nerable relationship is episodically
types in the water are taxonomicallyand reproductive rates (probably vary- disrupted by relatively rare extreme
the same as those in the symbiotic ing with host species or microhabitat events.
population, odds are high that a local-
within host). Elimination through alga-
ized/mild bleaching will produce alga competition is known among Research needs
reinfection with the type of zooxan-chlorellae (Rahat 1985), and Fitt
thella the host had before. Alterna-(1985) showed such elimination for There is increasing interest in attempt-
tively, reinfection by the same typezooxanthellae: in clams, those of a ing to define coral and reef commu-
may be due to "sufficient cells gener-
faster-growing type (isolated from a nity responses to rising temperature
ally remain[ing] for the coral to re-
sea anemone) gradually overwhelmed and other environmental stresses. Ex-
cover pigmentation after the stress is
algae of a slower-growing type (iso- perimentally induced bleaching and
removed" (Jokiel and Coles 1990, p. lated from a zoanthid). However, predictions of stress response are in-
156). The more severe the bleaching,
subtle differences in the microhabi- conclusive in the absence of definitive
the less likely the latter mechanismtats of various parts of a host might taxonomy of both symbiotic partners
and the greater the variety of symbi-
favor different types of algae, result- and some knowledge about the distri-
otic taxa likely to be affected, so the
ing in a mosaic or chimera, an effect butions, abundances, and diversity of
more likely corals are to be repopulated
that would be absent in controlled taxa or types in nature. It is important
by algae from the external environ- laboratory experiments. (In a chimera, to resolve these taxonomic and eco-
ment (Hayes and Bush 1990), result- what stresses one type of alga might logical issues before committing ex-
ing in a higher probability of change.
not stress another, allowing the resis- tensive resources to field or labora-
Repopulation is unlikely to be en-
tant one to remain and dominate the tory studies of inadequately
tirely random. Some combinations of host after bleaching.) characterized systems.
host and symbiont are not viable (Fitt We consider partial, background The hypothesis we advance bears
1985, Schoenberg and Trench 1980b).bleaching to be not pathological, but directly on these issues and can be
Thus, the original combination- rather a normal regulatory process verified or refuted, in part or in the
which, ipso facto, is viable-has a that maintains remarkably stable whole, by technically feasible experi-
higher-than-random probability of populations of symbiotic algae de- ments. Regardless of outcome, the
being re-established, particularly if
spite reproductive potential greatly in research we suggest would provide a
conditions return to prestress levels.
excess of that of their hosts (e.g., conceptually unified approach to a
However, to the extent that the Muscatine et al. 1985). It may also multifaceted subject.
alga is rendered incapable of infection
provide an important route of entry Symbiont specificity within indi-
by bleaching or that the stress persists,
for zooxanthellae under nonstress vidual coral colonies and in different
repopulation by a different type is conditions. This mechanism permits colonies of the same host species should
more likely. The factors that governcontinuous sampling of ambient al- be investigated in a range of sensitive
the types of zooxanthellae in an area
gae, the fittest ones of which can and robust symbiotic units, as well as
may be analogous to the factors spread to areas already populated by in diverse geographic locations.
thought to govern the diversity of algae that are either identical or less Within-colony studies should focus
corals on a reef (Oliver 1985): weedy
competitive. on corals with characteristics possibly
species tend to dominate in the short We infer from empirical evidence indicative of internal algal diversity-
term but their vulnerability to stress
of repetitive bleaching that algae in a for example, recent or repeated par-
eliminates them disproportionately, stress-resistant symbiotic unit may be tial bleaching, variations in pigmen-
checking their number and areal ex- gradually displaced by a less stress- tation or light exposure, and multiple
tent and allowing survival of those resistant type of zooxanthella that is, growth forms (e.g., a massive colony
that grow more slowly but are more in at least some symbiotic associa- with a platy skirt). Between-colony
stress resistant. tions, better adapted in other ways. comparisons should pursue local host
Indeed, Fitt (1985) found that hosts
For example, combinations with the differences in habitat, timing of re-
of faster-growing algal types grow highest individual or combined growth production, seasonality of skeletal
faster themselves; if rapid growth is
rate may not necessarily be the fittest density band formation (Lough and
associated with vulnerability to stress,
in all respects: anemones "infected Barnes 1990), depth, colony morphol-
then fast-growing types would readily
with homologous algae achieve a ogy and growth rate (Barnes and
establish but be subject to repeatedsomewhat larger biomass but pro- Lough 1989), pigmentation, or em-
bleaching. Reinfection, whether by duce fewer offspring than polyps in- pirically observed stress resistance. If
the original or by another type (or fected with [heterologous] zooxan- algal symbiont type can influence skel-
types) of alga, may be facilitated by
thellae" (Kinzie and Chee 1979, p. etal architecture, taxonomic problems
the attraction of motile zooxanthellae
324). Reversion to the symbiotic unit in genera where species are difficult to
to an aposymbiotic host (a normally existing before bleaching is most likely separate conclusively on the basis of
symbiotic individual lacking algae) if large populations o:fthe host exist, skeletal morphology (e.g., Knowlton

May 1993 323

This content downloaded from 66.194.72.152 on Sun, 15 Sep 2013 02:28:36 AM

All use subject to JSTOR Terms and Conditions
 of biochemical products (e.g., num-
ALGA-CONTROLLED PROCESSES
Infectionof host ber, type, condition, or effluvia of
times
Algallife/reproduction
Competitiveexclusionof otheralgae expelled zooxanthellae) that may act
Recoveryfromstress-inducedbleaching above some threshold as a signal analo-
gous to the alarm pheromone antho-
HOSTCHARACTERISTICS
Reproductionfrequencies
pleurine (Howe and Sheikh 1975). A
Organismlifespans mechanism for propagating mass
Reef community'lifespans"
ages, host families
Evolutionary bleachings beyond the effect of a ma-
age, scleractinianlineage
Evolutionary jor primary stress (e.g., to greater
ENVIRONMENTAL FACTORS
depth or protected sites) would help
Bleaching-inducingstress events, duration to explain the puzzling local distribu-
Bleaching-inducingstress events, retumperiod
Minorclimateoscillations
tion of some bleaching events. If the
Majorclimateoscillations algae precipitate the bleaching, they
could benefit by increasing the diver-
-2 -1
-6 -5 -4 -3 0 1 2 3
Log time (years)
4 5 6 7 8 9
sity or number of available hosts into
which to settle. Alternatively, if the
hosts were responsible for eliminating
Figure 1. Time scales relevant to bleaching-inducedadaptation. Bleaching and their algae, they could benefit by in-
infection mechanismsoperate at time scales characteristicof the motile and short- creasing the diversity or density of
lived algal symbionts.These time scales are short comparedwith host organismtime potential symbionts in the water.
constants and most environmentalvariation.The symbiotic unit is therebyable to Chemosensory attraction has been
adapt to stress and changes over time scales both longer and shorter than host implicated in the infection process
lifetimes.The punctuatedadaptationproducedby reshufflingsymbiontsresultsin a
(Fitt 1984).
highlevelof apparentevolutionarystabilityfor a relativelyspecializedand apparently
fragile organism.Dashed lines indicate possible but not probableextensions of the
ranges. Evolutionary and
ecological implications
et al. 1992) may be easier to under- scleractinian corals may acquire zoo- Except for the Merulinidae, all cur-
stand. Because the biogeography and xanthellae from host organisms of rently extant families of reef-forming
ecology of zooxanthella types is es- other taxa. Some symbiotic combina- corals had evolved by the Miocene
sentially unknown, it would be desir- tions may be obligate, whereas the (and most are Mesozoic in age; Wells
able to replicate these studies over physiological bond of others may be 1956). Most corals have lifetimes on
large latitudinal and longitudinal looser. Determination of preferences/ the order of decades (centuries in the
ranges. hierarchies will complement informa- case of some species), but during the
Surveys to determine the density, tion on the extent to which displace- Quaternary, reef environments have
variety, and infectivity of swarmers ment of one algal type by another repeatedly undergone rapid climate-
should be done in various localities occurs. induced environmental changes on
that differ in dominant symbiotic or- Empirical studies of stability and time scales of centuries to millennia
ganisms, and surveys should also be physiological features of various coral- (Figure 1). Under these conditions,
done through time (e.g., seasonally) alga partnerships should be accompa- the fact that the coral lineages have
at those same sites. It would be espe- nied by laboratory experiments on experienced neither rapid taxonomic
cially instructive to compare free-liv- variability in stress responses. These diversification nor substantial extinc-
ing zooxanthellae populations during experiments should address both natu- tions (Wells 1956) is perhaps unex-
or immediately after a bleaching epi- rally occurring symbiotic combina- pected, and the reasons may prove re-
sode with those during normal cir- tions and deliberate cross-infections levant to present concerns about the
cumstances. and should investigate variability in effects of greenhouse effect-induced
Laboratory studies on stress re- competitive interactions among algal climate change (Smithand Buddemeier
sponses (especially to temperature) of types under various conditions, ex- 1992).
various types of free-living zooxan- panding on work such as that by The rapid potential generation time
thellae are necessary to test our hy- Kinzie and Chee (1979), Fitt and of zooxanthellae compared with that
pothesis. These studies should investi- Trench (1981), and Fitt (1985). of most hosts might appear to make
gate not only viability but also effects Our hypothesis has implications them (and hence their symbiotic units)
on reproduction, photosynthesis and for applied research. It suggests that evolutionarily more responsive to lo-
respiration, infectivity, and biochemi- ecosystem management and restora- cal conditions than the hosts. How-
cal products. tion could become simpler and more ever, a sexual phase is unknown for
We need to know the hierarchy of effective if resistance to stress can be zooxanthellae (Blank and Trench
host species preference by various enhanced by manipulating popula- 1985b), which calls into question their
zooxanthella types and vice versa, tions of algae instead of the larger and potential evolutionary responsiveness.
and how this hierarchy may vary with slower-growing hosts. The probable evolutionary stasis of
environmental parameters. Studies Finally, although not integral to both host and symbiont may be com-
must include the complete spectrum of our hypothesis, we suggest testing the pensated for by the creation of new
reef-dwelling symbiotic units, because possibility that stress induces release symbiotic units through recombining

324 BioScience Vol. 43 No. 5

This content downloaded from 66.194.72.152 on Sun, 15 Sep 2013 02:28:36 AM

All use subject to JSTOR Terms and Conditions
symbiotic partners.
We view bleaching as providing an
opportunity for reshuffling, a potent
adaptive mechanism that instantly
creates a host-symbiont combination
with features that may prove more
robust under altered conditions. Such
a mechanism of adaptation, which
Saffo (1992) considered as indirectly
affecting evolution, could account not
only for currently observed local and
regional variations in sensitivity to
stress of nominally identical host spe-
cies, but also for the remarkable ro-
bustness of coral species and reef
community structure throughout the
climatic oscillations of the Quaternary.
Time scales of some environmental
variations are comparable to those of
background cross-infection and com-
petitive elimination of algae from a
host (Figure 1). Therefore, the hetero-
geneity of algae in a host at any time
reflects some combination of com-
petitiveness of the algae, time since
the most recent stress, frequency of
stress, and local diversity of zooxan-
thellae. Rapid environmental change
may thereby produce punctuated ad-
aptation of the symbiotic units, which
has been neglected as a subject of
study because of the tendency to as-
cribe stable attributes to the host spe-
cies as a unitary entity.
Rather than an absence of specific-
ity between partners, our hypothesis
posits that specificity changes as envi-
ronmental conditions do. Thus, al-
though bleaching may represent in-
stability in the short term, it promotes
long-term stability by enhancing sur-
vival chances of both zooxanthellae
and hosts under conditions that are
not those of the prestress environ-
ment. Potential hosts and habitats are
many, so the combination of environ-
mental variability and symbiotic ad-
aptation acts to conserve both genetic
and symbiotic diversity.
Acknowledgments
We are grateful to colleagues too nu-
merous to mention for the time and
attention they gave to a previous ver-
sion of this manuscript. Their com-
ments helped us to focus our ideas and
sharpen our prose.
References cited
Barnes, D. J., and J. M. Lough. 1989. The
nature of skeletal density banding in
May 1993
This content downloaded from 66.194.72.152 on Sun, 15 Sep 2013 02:28:36 AM
scleractinian corals: fine banding and sea-
sonal patterns.J. Exp. Mar. Biol. Ecol. 126:
119-134.
Berner, T., Y. Achituv, Z. Dubinsky, and Y.
Benayahu. 1987. Pattern of distribution and
adaptation to different irradiance levels of
zooxanthellae in the soft coral Litophyton
arboreum (Octocorallia, Alcyonacea). Sym-
biosis 3: 23-40.
Blank, R. J. 1987. Evolutionary differentiation
in gymnodinoid zooxanthellae. Ann. N.Y.
Acad. Sci. 503: 530-533.
Blank, R. J., and R. K. Trench. 1985a. Specia-
tion and symbiotic dinoflagellates. Science
229:656-658.
.1985b. Symbiodinium microadriati-
cum: a single species? Proc. Intl. Coral Reef
Congr. 6:113-117.
Brown, B. E., ed. 1990. Coral bleaching: special
issue. Coral Reefs 8: 153-232.
Brown, B. E., and Suharsono. 1990. Damage
and recovery of coral reefs affected by El
Nifio related seawater warming in the Thou-
sand Islands, Indonesia. Coral Reefs 8:
163-170.
Ciereszko, L. S. 1962. Chemistry of coelenter-
ates. III. Occurrence of antimicrobial terpe-
noid compounds in the zooxanthellae of
alcyonarians. Trans. N.Y. Acad. Sci. 24:
502-503.
.1991. Sterol and diterpenoid produc-
tion by zooxanthellae in coral reefs: a re-
view. Biol. Oceanogr. 6: 363-374.
Coles, S. L., and P. L. Jokiei. 1978. Synergistic
effects of temperature, salinity and light on
the hermatypic coral Montipora verrucosa.
Mar. Biol. 49: 187-195.
Coles, S. L., P. L. Jokiel, and C. R. Lewis. 1976.
Thermal tolerance in tropical versus sub-
tropical Pacific reef corals. Pac. Sci. 30:
159-166.
Cook, C. B., A. Logan, J. Ward, B. Luckhurst,
and C. J. Berg Jr. 1990. Elevated tempera-
tures and bleaching on a high-latitude coral
reef: the 1988 Bermuda event. Coral Reefs
9:45-49.
D'Elia, C. F., R. W. Buddemeier, and S. V.
Smith. 1991. Workshop on coral bleaching,
coral reef ecosystems and global change:
report of proceedings. Maryland Sea Grant
College publication no. UM-SG-TS-91-03,
University of Maryland, College Park.
Dustan, P. 1982. Depth-dependent photo-
adaptation by zooxanthellae of the reef
coral Montastrea annularis. Mar. Biol. 68:
253-264.
Fautin, D. G. 1991. The anemonefish symbio-
sis: what is known and what is not. Symbio-
sis 10: 23-46.
Fisk, D. A., and T. J. Done. 1985. Taxonomic
and bathymetric patterns of bleaching in
corals, Myrimidon Reef (Queensland). Proc.
Intl. Coral Reef Congr. 6: 149-154.
Fitt, W. K. 1984. The role of chemosensory
behavior of Symbiodinium microadriaticum,
intermediate hosts, and host behavior in the
infection of coelenterates and molluscs with
zooxanthellae. Mar. Biol. 81: 9-17.
.1985. Effect of different strains of the
zooxanthella Symbiodinium micro-
adriaticum on growth and survival of their
coelenterate and molluscan hosts. Proc. Intl.
Coral Reef Congr. 6: 131-136.
Fitt, W. K., and R. K. Trench. 1981. Spawning,
development,and acquisition of zooxan-
thellae by Tridacna squamosa (Mollusca,
All use subject to JSTOR Terms and Conditions
Bivalvia). Biol. Bull. 161: 213-235.
Freudenthal, H. D. 1962. Symbiodinium gen.
nov. and S. microadriaticum sp. nov., a
zooxanthella: taxonomy, life cycle and
morphology. J. Protozool. 9: 45-52.
Gates, R. 1990. Seawater temperature and
sublethal coral bleaching in Jamaica. Coral
Reefs 8: 193-197.
Gates, R. D., G. Baghdasarian, and L.
Muscatine. 1992. Temperature stress causes
host cell detachment in symbiotic cnidarians:
implications for coral bleaching. Biol. Bull.
182:324-332.
Glynn, P. W. 1990. Coral mortality and distur-
bances to coral reefs in the tropical Eastern
Pacific. Pages 55-126 in P. W. Glynn, ed.
Global Ecological Consequences of the
1982-1983 El Nino-Southern Oscillation.
Elsevier, New York.
.1991. Coral reef bleaching in the 1980's
and possible connections with global warm-
ing. Trends Ecol. Evol. 6: 175-179.
Glynn, P. W. and W. H. de Weert. 1991. Elimi-
nation of two reef-building hydrocorals fol-
lowing the 1982-83 El Nifio warm event.
Science 253: 69-71.
Glynn, P. W., and J. S. Feingold. 1992. Hydro-
coral species not extinct. Science 257:1845.
Goreau, T. J. 1990. Coral bleaching in Jamaica.
Nature 343:417.
Hayes, R. L., and P. G. Bush. 1990. Micro-
scopic observations of recovery in the reef-
building scleractinian coral, Montastrea
annularis, after bleaching on a Cayman
reef. Coral Reefs 8: 203-209.
Hoeksema, B. W. 1991. Control of bleaching in
mushroom coral populations (Scleractinia:
Fungiidae) in the Java Sea: stress tolerance
and interference by life history strategy.
Mar. Biol. Prog. Ser. 74: 225-237.
Howe, N. R., and Y. M. Sheikh. 1975.
Anthopleurine: a sea anemone alarm phero-
mone. Science 189: 386-388.
Jaap, W. C. 1985. An epidemic zooxanthellae
expulsion during 1983 in the lower Florida
Keys coral reefs: hyperthermic etiology.
Proc. Intl. Coral Reef Congr. 6: 143-148.
Jokiel, P. L., and S. L. Coles. 1990. Response of
Hawaiian and other Indo-Pacific reef corals
to elevated temperature. Coral Reefs 8:
155-162.
Karakashian, S. J., and R. W. Siegel. 1965. A
genetic approach to endocellular symbiosis.
Exp. Parasitol. 17: 103-122.
Kinzie, R. A. III, and G. S. Chee. 1979. The
effect of different zooxanthellae on the
growth of experimentally reinfected hosts.
Biol. Bull. 156: 315-327.
Knowlton, N., E. Weil, L. A. Weigt, and H. M.
Guzman. 1992. Sibling species in Mon-
tastraea annularis, coral bleaching, and the
coral climate record. Science 255: 330-333.
Lang,J. C., R. I. Wicklund, and R. F. Dill. 1988.
Depth- and habitat-related bleaching of
zooxanthellate reef organisms near Lee
Stocking Island, Exuma Cays, Bahamas.
Proc. Intl. Coral Reef Symp. 3: 269-274.
Lough, J. M., and D. J. Barnes. 1990. Intra-
annual timing of density band formation of
Porites coral from the central Great Barrier
Reef. J. Exp. Mar. Biol. Ecol. 135: 35-37.
Muller Parker, G. 1984. Dispersal of zooxan-
thellae on coral reefs by predators of
cnidarians. Biol. Bull. 167: 159-167.
Muscatine, L. 1971. Experiments on green al-
gae coexistent with zooxanthellae in sea
325

Fish
An Enthusiast's
Guide
PETERB. MOYLE
byChrisMar vanDyck
Illustrated
Moyle introducesus to the great
advancesin knowledgeof fish
biologyand behaviorthathave
occurred over the last twenty
years.VanDyck'sdramatic paint-
ingsandexquisitelinedrawings
illustrateMoyle'sdescriptions.
$25.00 cloth,illustrated
Thelepson
Manual
HigherPlants
of California
JAMESC. HICKMAN,Editor
"A single work . . . simultan-
eously accessibleto dedicated
beginnersand indispensableto
professional botanists.... For
the first time in one volume a
floraof theexceed-
user-friendly
ingly diverse higher plants of
California."
-Mildred E. Mathias,editorof
Flowering Plantsin theLandscape
"Setsnew standardsfor excel-
lence... andpicksup beautifully
on the contemporaryidea that
botanicalwork should be fully
accessibleto the generalpublic
as well as to scientists."
-Peter H. Raven,Missouri
BotanicalGarden
$55.00 clothuntil6/30/93,
$65.00 after,illustrated
At bookstoresor ordertoll-free1-800-822-6657.
Universityof CaliforniaPress *
BerkeleyLosAngeles New York London
This content downloaded from 66.194.72.152 on Sun, 15 Sep 2013 02:28:36 AM
anemones. Pac. Sci. 25: 13-21.
Muscatine, L., L. R. McCloskey, and Y. Loya.
1985. A comparison of the growth rates of
zooxanthellae and animal tissue in the Red
Sea coral Stylophora pistillata. Proc. Intl.
Coral Reef Congr. 6: 119-123.
LatinAmerican Oliver, J. 1985. Recurrent seasonal bleaching
and mortality of corals on the Great Barrier
Insects Reef. Proc. Intl. Coral Reef Congr. 4:
201-206.
and Entomology Rahat, M. 1985. Competition between
Chlorellae in chimeric infections of Hydra
CHARLESL. HOGUE viridis: the evolution of a stable symbiosis.
This is the firstcomprehensive J. Cell. Sci. 77: 87-92.
guideto insectlifein a partof the Rowan, R., and D. A. Powers. 1991. A molecu-
world known for its abundant, lar genetic classification of zooxanthellae
and the evolution of animal-algal symbio-
and endangered, life forms. ses. Science 251: 1348-1351.
Hogue's scholarshipembraces Saffo, M. B. 1992. Invertebrates in endosymbi-
vast geographicalterritoryand otic associations. Am. Zool. 32: 557-565.
Salvat, B. 1991. Bleaching and mortality of
generousnumbersof represen- scleractinian corals, event and causes, Soci-
tativeinsects,to serveboth the ety Islands, 1991. Paper presented at the
annual meeting of the International Society
professionalentomologistand for Reef Studies, Berkeley, CA, 14 Decem-
the seriousamateur. ber 1991.
$85.00 cloth,illustrated Schoenberg, D. A., and R. K. Trench. 1976.
Specificity of symbioses between marine
cnidarians and zooxanthellae. Pages
423-432 in G. O. Mackie, ed. Coelenterate
Ecology and Behavior. Plenum Press, New
York.
1980a. Genetic variation in Symbio-
dinium (=Gymnodinium) microadriaticum
Freudenthal, and specificity in its symbiosis
with marine invertebrates. I. Isozyme and
I
i ~~~~~
4
I
v
t
soluble protein patterns of axenic cultures
of Symbiodinium microadriaticum. Proc.
R. Soc. Lond. B 207: 405-427.
.1980b. Genetic variation in Symbiodi-
nium (=Gymnodinium) microadriaticum
Freudenthal, and specificity in its symbiosis
with marine invertebrates. III. Specificity
and infectivity of Symbiodinium microa-
driaticum. Proc. R. Soc. Lond. B 207:
AnnualCycles 445-460.
Smith, S. V., and R. W. Buddemeier. 1992.
of Diving Global change and coral reef ecosystems.
Annu. Rev. Ecol. Syst. 23: 89-118.
Behaviorand Trench, R. K. 1971. The physiology and bio-
chemistry of zooxanthellae symbiotic with
Ecologyof the marine coelenterates. II. Liberation of fixed
14Cby zooxanthellae in vitro. Proc. R. Soc.
WeddellSeal Lond. B 177: 237-250.
Von Holt, C., and M. Von Holt. 1968. The
MICHAELA. CASTELLINI, secretion of organic compounds by zooxan-
RANDALLW. DAVIS,and thellae isolated from various types of
Zoanthus. Comp. Biochem. Physiol. 24:
GERALDL. KOOYMEN 83-92.
This work details more than Wells, J. W. 1956. Scleractinia. Pages
ten years of work in the Ant- F328-F444 in R. C. Moore, ed. Treatise on
arcticon the diving biology of Invertebrate Paleontology. Part F.
Coelenterata. Geological Society of America
Weddell seals. and University of Kansas Press, Lawrence.
Bulletinof theScrippsInstitution
of Williams, E. H. Jr., and L. Bunkley-Williams.
Oceanography 1990. The world-wide coral reef bleaching
$8.00 paper
cycle and related sources of coral mortality.
Atoll-Res. Bull. 335: 1-71.
1893 1 9 93
BioScience Vol. 43 No. 5
All use subject to JSTOR Terms and Conditions